• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제6권1호
• /
• pp.10-14
• /
• 2006

Many clustering methods have been studied. For the most part of these methods may be needed to determine the number of clusters. But, there are few methods for determining the number of population clusters objectively. It is difficult to determine the cluster size. In general, the number of clusters is decided by subjectively prior knowledge. Because the results of clustering depend on the number of clusters, it must be determined seriously. In this paper, we propose an efficient method for determining the number of clusters using hybrid' self organizing map and new criterion for evaluating the clustering result. In the experiment, we verify our model to compare other clustering methods using the data sets from UCI machine learning repository.

• 공학교육연구
• /
• 제16권3호
• /
• pp.10-19
• /
• 2013

Recently, Korean universities have increased the number of English Medium Instruction (EMI) lectures in order to allow students to gain both specialized knowledge and enhanced English ability. Previous researches on effective EMI lectures have focused on exploring the effects of learners' cognitive and affective characteristics on learning outcomes. Whereas the input variables of learning have been investigated as predicting variables of effects in EMI lectures, there has been a few research for investigating process variables to yield learning outcomes. The purpose of this study is to analyze the structural relationships among variables affecting on learner satisfaction and effects. The participants are 209 engineering students from various majors. Independent variables are defined as English motivation, English competency, and English confidence, a mediated variable is Cognitive engagement, and dependent variables are Learning satisfaction and Educational effect perception. The results show that the relationships are statistically significant: learners' English competency & English confidence ${\rightarrow}$ Cognitive engagement ${\rightarrow}$ Learning satisfaction ${\rightarrow}$ Educational effect perception. Especially, the structural model confirms that the effect of learners' English confidence on Learning satisfaction and Educational effect perception is mediated by the level of learners' Cognitive engagement. Further, the implication for effective EMI lectures is discussed based on the observed research results.

• 한국응급구조학회지
• /
• 제20권3호
• /
• pp.21-35
• /
• 2016

Purpose: The purpose of the study was to investigate the continuous effect of advanced cardiovascular life support (ACLS) simulation education according to Felder-Silverman learning style. Methods: A self-reported questionnaire was completed by 94 students of emergency medical technology and nursing. There were 50 female students (53.2%) and 88 students (93.6%) had basic life support certification. The study instruments included knowledge, performance, and confidence. Data were analyzed using SPSS v. 20.0. Results: The learning style consisted of reflective type (51.1%), sensory type (76.6%), visual type (63.8%), and sequential type (64.9%). There was a significant difference in continuous effect on performance by learning type. Conclusion: It is necessary to identify the learning style of students before simulation education in order to maintain continuous effect of ACLS education.

• 한국도서관정보학회지
• /
• 제41권4호
• /
• pp.355-381
• /
• 2010

사회 환경 측면이나 도서관 현장에서 독서프로그램 담당 사서의 수요가 지속적으로 요구됨에도 불구하고 문헌정보학과에서는 독서관련 교과목 개설이 미비하며, 교육내용 또한 이론위주의 교육이 이루어지고 있다. 이에 본 연구에서는 독서교육자의 자질을 함양하기 위한 실무중심의 독서관련 교과목 교수-학습법을 개발하였다. 즉, 독서체험을 위한 자기주도적 독서 및 독후포트폴리오 제작활동과 실무 중심의 학습동기 유발 독서프로그램 운영과 도서관현장 연계 협동 프로젝트 수행에 대한 교수-학습법을 개발 적용하였다.

• 한국멀티미디어학회논문지
• /
• 제23권2호
• /
• pp.135-145
• /
• 2020

Recently, deep-learning based methods for low-light image enhancement accomplish great success through supervised learning. However, they still suffer from the lack of sufficient training data due to difficulty of obtaining a large amount of low-/normal-light image pairs in real environments. In this paper, we propose an unsupervised learning approach for single low-light image enhancement using the bright channel prior (BCP), which gives the constraint that the brightest pixel in a small patch is likely to be close to 1. With this prior, pseudo ground-truth is first generated to establish an unsupervised loss function. The proposed enhancement network is then trained using the proposed unsupervised loss function. To the best of our knowledge, this is the first attempt that performs a low-light image enhancement through unsupervised learning. In addition, we introduce a self-attention map for preserving image details and naturalness in the enhanced result. We validate the proposed method on various public datasets, demonstrating that our method achieves competitive performance over state-of-the-arts.

• 대한지구과학교육학회지
• /
• 제3권3호
• /
• pp.163-174
• /
• 2010

The purpose of this study is to device the teaching-learning method and the evaluating method on 'free inquiry' which is newly introduced in 2007 revised curriculum. The teaching-learning models which apply 'free inquiry' are 'group investigation', 'PBL', 'project investigation' and 'IIM'. And the unit which apply 'free inquiry' on the fourth grade and the second semester in the elementary science. The results of this study are as follows: First, The person leading the lesson is not a teacher but learners. The focus of teaching-learning is not a unit but topic, problem or project on the science textbook. A teacher's role is not the deliverer of knowledge but the guide of learning. Second, the outcome of applying group investigation, PBL project investigation and IIM to 'free inquiry' is improvements of the problem-solving ability and the self-directed learning ability as well as building scientific attitude and social skills as educational effect in commonly. Third, to apply 'free inquiry' efficiently, teachers should understand each subject very well, teach a class with a thorough and concrete plan, and try to evaluate objectively.

• East Asian mathematical journal
• /
• 제36권2호
• /
• pp.229-251
• /
• 2020

Problem Based learning(PBL) is a teaching and learning method to increase mathematical ability and help achieving mathematical concepts and principles through problem solving using the learner's mathematical prerequisite knowledge. In addition, the recent instructional situations or environments have focused on the learner's self construction of his learning and its process. In spite of such a quite attention, it is not easy to apply and execute PBL program actually in class. Especially, there are some difficulties in actually applying and practicing PBL in the areas of mathematics education in not only secondary school but also in college. Its reason is that in order to conduct PBL instruction constantly in real or experimental class there is no more concrete and detailed instructional content during the consistent and long period. However, to whom is related to mathematics education including instructors called scaffolders, investigation and recognition on the degree of the learner's acquisition of mathematical thinking skills and strategies is an very important work. By the reason, in this study, the instructional content was to be explored and developed to be conducted during 15 weeks in one semester, which was based on Problem Based Learning environment by the subject of the theories relevant to mathematics education in the college of education.

• 지식경영연구
• /
• 제17권2호
• /
• pp.27-50
• /
• 2016

Korean manufacturing industry have recently faced the catch-up of China in the mass commodity product, such as automotive, display, and smart phone in terms of market as well as technology. Accordingly, discussion on the importance of achieving catch-up in complex product systems (CoPS) has been increasing as a new innovation engine for the industry. In order to achieve successful catch-up of CoPS, we explored emerging technologies of CoPS, which are featured by the characteristics of radical novelty, relatively fast growth and self-sustaining, through the study of emerging technologies of gas turbine for power generation. We found that emerging technologies of the gas turbine are technologies for combustion nozzle and composition of electrical machine for increasing power efficiency, washing technology for particulate matter, cast and material processing technology for enhancing durability from fatigue, cooling technologies from extremely high temperature, interconnection operation technology between renewable energy and the gas turbine for flexibility in power generation, and big data technology for remote monitoring and diagnosis of the gas turbine. We also found that those emerging technologies resulted in technological progress of the gas turbine by converging with other conventional technologies in the gas turbine. It indicates that emerging technologies in CoPS can be appeared on various technological knowledge fields and have complementary relationship with conventional technologies for technology progress of CoPS. It also implies that latecomers need to pursue integrated learning that includes emerging technologies as well as conventional technologies rather than independent learning related to emerging technologies for successful catch-up of CoPS. Our findings provide an important initial theoretical ground for investigating the emerging technologies and their characteristics in CoPS as well as recognizing knowledge management strategy for successful catch-up of latecomers. Our findings also contribute to the policy development of the CoPS from the perspective of innovation strategy and knowledge management.

• 부모자녀건강학회지
• /
• 제12권1호
• /
• pp.77-87
• /
• 2009

Purpose: The purpose of this study was to determine the knowledge of nocturnal emissions, or wet dreams and to provide information for developing a wet dream education program for elementary school boys. Method: The research design was a descriptive study through questionnaires. Data was collected from May 10, 2009 to May 20, 2009. A total of 197 late elementary school boys in Incheon, Korea participated in this study. Data was analyzed using the SPSS WIN 15.0 program. Result: The results of this study were as follows; Knowledge of wet dreams in late elementary school boys was low; The mean of the information about wet dreams was 14.00 on a 100 point scale; There were significant differences of knowledge regarding wet dreams according to the educational experience at 6 items and wet dreams experience at 12 items of total 21 items. This indicates that present education for wet dreams is not efficient and does not provide appropriate information when they need practical information. Conclusion: Self education learning program for wet dreams should be developed for late elementary school boys which includes different contents and time required for further study, evaluation and assessment.

• Communications for Statistical Applications and Methods
• /
• 제30권5호
• /
• pp.501-516
• /
• 2023

Knowledge distillation (KD) refers to extracting knowledge from a large and complex model (teacher) and transferring it to a relatively small model (student). This can be done by training the teacher model to obtain the activation function values of the hidden or the output layers and then retraining the student model using the same training data with the obtained values. Recently, relational KD (RKD) has been proposed to extract knowledge about relative differences in training data. This method improved the performance of the student model compared to conventional KDs. In this paper, we propose a new method for RKD by introducing a new loss function for RKD. The proposed loss function is defined using the area difference between the teacher model and the student model in a specific hidden layer, and it is shown that the model can be successfully compressed, and the generalization performance of the model can be improved. We demonstrate that the accuracy of the model applying the method proposed in the study of model compression of audio data is up to 1.8% higher than that of the existing method. For the study of model generalization, we demonstrate that the model has up to 0.5% better performance in accuracy when introducing the RKD method to self-KD using image data.